Post on 02-Feb-2021
Monitoring dissolved organic matterusing submersible tryptophan-like
fluorometers
Kieran Khamis1,2, J. Sorensen3, C. Bradley2, D. Hannah2, R. Stevens1
1 RS Hydro2 School of Geog. Earth & Env .Sci. University of Birmingham3 British Geological Survey
•Fluorescence: a form of luminescence which occursover short time scales at the molecular/atomic level.
What is fluorescence?
Sjoback et al. (1995) Spec. Acta A 51, 7-21
How does it work?
• The fluorescent spectra of compounds important forwater quality monitoring have been identified
ProteinsChlorophyll‘Humic’ matter
Natural fluorescent compounds
Excitation wavelength
Emis
sion
wav
elen
gth
EEM spectroscopy
Excitation Emission Matrix (EEM)
Bench top scanning fluorometer
Humic-like compounds(terrestrial origin)
Protein-like compounds(in stream) – indicativeof organic enrichment
However…..Not suitable for remote field sites or if high resolution records are required.
Tryptophan - like peak related tomicrobial activity + correlated withBOD5
Hudson et al. (2008) SOTE. 391, 149-158Fellman et al. (2012) Lim. & Oce. 55, 2452
Submersible fluorometers
• Quenching – e.g. temperature;
• Matrix interference – e.g. suspended particles inwater column;
• Inner-filtering - concentration effect;
• Measurement repeatability - between/within sitesand between sensors;
• To date no rigorous tests of submersible tryptophanfluorometers have been conducted.
Challenges to in-situ monitoring
Baker (2005) Water Research 39, 4405
Downing et al. (2012) Lim. Oce. Methods 10, 767
Ohno (2002) Env. Sci. Tech. 36, 742
Watras et al. (2011) Lim. Oce. Methods 9, 296
Study objectives
The objectives of this study were to:
1. Test the performance of two commerciallyavailable tryptophan fluorometers in thelab;
2. Develop empirical correction factors toaccount for fluorescence quenching andmatrix interference;
3. Undertake a field trial to assess sensorperformance and test correction factors.
Minimum Detection Limit (MDL) and precision
T1 T2 C1 C2
Calibrated relationship y = 0.997x - 0.133 y = 1x + 0.0009 y = 1x - 0.00007 y = 1x + 0.00006
Relationship with Varian (ppb) y = 0.99x - 0.1255 Y = 1x + 0.0022 y = 1x + 0.0076 y = 0.99x + 0.0129
Relationship with Varian (R.U) y = 0.002x + 0.0041 y = 0.002x + 0.0044 y = 0.002x + 0.0044 y = 0.002x + 0.0044
MDL ± SD 1.99 ± 0.53 1.92 ± 0.57 0.17 ± 0.06 0.19 ± 0.15
Precision: CV (5ppb) 3.03 2.49 0.45 0.22
(50ppb) 0.03 0.02 0.00 0.01
(400ppb) 3.79-4 4.86-4 4.63-4 6.27-4
Accuracy (RMSE) 0.63 0.62 0.57 0.58
Linearity
C1 C2 T1 T2Significant difference inprecision at low concentration
Thermal quenching
Chelsea (C2)
100 ppb
50 ppb
25 ppb10 ppb
Chelsea (C2)
Ratio correction Exp. correction
RSR = 0.2
RSR = 0.51
RSR = 0.35
RSR = 0.41
RSR = 0.36
RSR = 0.15
=
Raw data
Turbidity interference
100 NTU
900 NTU200 NTU
Turbidity interference (Clay)
• Particle size influencesresponse: attenuationgreater with clay.
• Sensor specificresponses: differencesbetween and withinmanufacturers.
• Sensor specificcalibrations may berequired
Turbidity interference(silt)
Out of range for 500ppb
Turbidity interference (clay)
95% CI overlap> 200 NTU 95% CI overlap
> 200 NTU
95% CI overlap> 200 NTU
Turbidity interference (silt)
95% CI overlap> 600 NTU
95% CI overlap> 800 NTU
No 95% CIoverlap
Turbidity correction
Clay (Fullers Earth) Silt (glacial outwash)
Urban field test site
Carstea et al. (2009) Hydrological Processes 23,1937
Chelsea fluorometer and Manta 2-Stage-Turbidity-EC-Tw
-Tryptophan
ISCO pump sampler
Urban field test site
Field trial
Hurricane BerthaEvent characterised
Field trial: raw data
Field trial: corrected data
T1
T1
C1
C1
FE = Fullers Earth (clay) GS = Glacial silt
Field trial: corrected data
R2= 0.92m = 0.69 ± 0.04c = 0.91 ± 4.53
R2= 0.92m = 0.74 ± 0.04c = 1.17 ± 4.55
R2= 0.88m = 0.95 ± 0.07c = -2.41 ± 5.90
R2= 0.91m = 0.67 ± 0.04c = 7.95 ± 4.38
T1 raw (ppb) T1 silt + Tw correction (ppb)
T1 Tw correction (ppb) T1 clay + Tw correction (ppb)
Field trial: corrected data
R2= 0.77m = 0.80 ± 0.09c = -23.0 ± 10.84
R2= 0.76m = 0.86 ± 0.10C = -22.4 ± 10.86
R2= 0.56m = 1.04 ± 0.18c = -14.71 ± 15.58
R2= 0.76m = 0.96 ± 0.11c = -2.03 ± 8.75
C1 raw (ppb) C1 silt + Tw correction (ppb)
C1 Tw correction (ppb) C1 clay + Tw correction (ppb)
Borehole test
Pond River
C1
C2
T1
T2
Lab (Varian)
Spatial survey (initial result)
All sites: R2 = 0.60
River sites: R2 = 0.91
Habitat type – clear residualpattern
Canal Pond River Effluent
River samples: R2 = 0.67
All samples: R2 = 0.72 River samples: R2 = 0.64
Conclusions
Quenching of T1 fluorescence was identified in the lab and varied betweensensors (Turner & Chelsea)
Temperature compensation appears relatively simple but evidence ofhysteresis requires further investigation
Sediment particle size influenced sensor response to turbidity increases(implies site specific calibrations may be necessary)
Field tests highlight the potential to develop and apply correction factors toimprove in-situ data output during both baseflow and event conditions
Further work will improve calibration for BOD5 - T1 fluorescence
Acknowledgments
NERC and EPSRC (co-funding project)
Les Basford (Nature centre, Birmingham)
Ed Lang and James Chapman (RS Hydro)
Alex Taylor (West Country Rivers Trust)
Richard Johnson and Mel Bickerton(University of Birmingham)
Pete Williams (BGS)